{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,29]],"date-time":"2026-07-29T06:59:20Z","timestamp":1785308360709,"version":"3.55.0"},"reference-count":36,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,7,29]],"date-time":"2026-07-29T00:00:00Z","timestamp":1785283200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>Adverse drug reactions (ADRs) are caused by medication and are considered a serious issue in healthcare when there is simultaneous use of different medications resulting in drug-drug interaction (DDI). Traditional approaches mostly focus on the effects caused by a single drug, and they fail to capture the side effects from drug combinations. In this research work, a deep learning-based DrugPred framework is proposed to predict ADR risks by integrating individual drug effects, interaction statistics of the drugs, and structural relationships between drug-interaction graphs. In the first phase, baseline association scores between individual drugs and their corresponding ADRs, such as nausea, liver toxicity, or cardiovascular effects, are learned using a multi-layer perceptron (MLP) trained on the OFFSIDES dataset, providing probabilistic values for individual drug-ADR pairs. In the second phase, a Graph Neural Network (GNN) based on the EdgeConv architecture is utilized to model drug-drug interactions using the TwoSIDES dataset, capturing relational dependencies by utilizing pairwise statistical values such as co-occurrence frequency and interaction statistics derived from pharmacovigilance signals, including transformed PRR-based signals. In the third phase, Bio_ClinicalBERT, which is a pre-trained transformer-based language model specifically trained on clinical notes and biomedical text, is used to encode drug pair representations and combine them with association and interaction scores through an attention-guided mechanism, which enables the model to adaptively weight heterogeneous features and finally performs multi-label ADR prediction. The proposed model achieves an accuracy of 95.73%, F1-score of 0.94, ROC-AUC of 0.993, and PR-AUC of 0.988. These results indicate that our proposed DrugPred framework is effective in the prediction of ADR risks with high precision and recall. Overall, the proposed framework provides a scalable approach for predicting ADRs in multi-drug settings. We further extend our framework with a retrieval-based guidance system that maps the risk levels of the predicted ADRs to appropriate System Organ Classes (SOC) and medical specialists to provide recommendations using real-world clinical data sources like PubMed and OpenFDA.<\/jats:p>","DOI":"10.3389\/frai.2026.1874825","type":"journal-article","created":{"date-parts":[[2026,7,29]],"date-time":"2026-07-29T06:02:20Z","timestamp":1785304940000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["DrugPred: an EdgeConv-GNN and Bio_ClinicalBERT based polypharmacy ADR prediction and specialist recommendation model"],"prefix":"10.3389","volume":"9","author":[{"given":"J. G.","family":"Arjay","sequence":"first","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Chennai, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"K. Adit","family":"Pushan","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Chennai, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"N.","family":"Roshanthraj","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Chennai, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Brindha","family":"Subburaj","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Chennai, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sherly","family":"Alphonse","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Chennai, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Girish","family":"Subramanian","sequence":"additional","affiliation":[{"name":"School of Business Administration, Penn State Harrisburg","place":["Middletown, PA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,7,29]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"2955","DOI":"10.1093\/bioinformatics\/btab174","article-title":"Polypharmacy side-effect prediction with enhanced interpretability based on graph feature attention network","volume":"37","author":"Bang","year":"2021","journal-title":"Bioinformatics"},{"key":"B2","doi-asserted-by":"publisher","first-page":"315","DOI":"10.1007\/s002280050466","article-title":"A Bayesian neural network method for adverse drug reaction signal generation","volume":"54","author":"Bate","year":"1998","journal-title":"Eur. J. Clin. Pharmacol"},{"key":"B3","unstructured":""},{"key":"B4","first-page":"2787","article-title":"\u201cTranslating embeddings for modeling multi-relational data\u201d","volume-title":"Advances in Neural Information Processing Systems (NIPS)","author":"Bordes","year":"2013"},{"key":"B5","doi-asserted-by":"publisher","first-page":"93823","DOI":"10.1109\/ACCESS.2024.3407877","article-title":"Detecting side effects of adverse drug reactions through drug-drug interactions using graph neural networks and self-supervised learning","volume":"12","author":"ChandraUmakantham","year":"2024","journal-title":"IEEE Access"},{"key":"B6","doi-asserted-by":"publisher","first-page":"831","DOI":"10.1109\/JBHI.2024.3492005","article-title":"ADR-DQPU: a novel ADR signal detection using deep reinforcement and positive-unlabeled learning","volume":"29","author":"Chung","year":"2025","journal-title":"IEEE J. Biomed. Health Inform"},{"key":"B7","doi-asserted-by":"publisher","first-page":"1791","DOI":"10.1109\/TCSS.2022.3231701","article-title":"Deep-learning-based drug recommendation and ADR detection healthcare model on social media","volume":"10","author":"Dongre","year":"2023","journal-title":"IEEE Trans. Comput. Soc. Syst"},{"key":"B8","doi-asserted-by":"publisher","first-page":"483","DOI":"10.1002\/pds.677","article-title":"Use of proportional reporting ratios (PRRs) for signal generation from spontaneous adverse drug reaction reports","volume":"10","author":"Evans","year":"2001","journal-title":"Pharmacoepidemiol. Drug Saf"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pbio.2005343","article-title":"Best match: new relevance search for PubMed","author":"Fiorini","year":"2018","journal-title":"PLoS Biol."},{"key":"B10","doi-asserted-by":"publisher","first-page":"1773","DOI":"10.1109\/JBHI.2024.3349570","article-title":"AutoDDI: drug\u2013drug interaction prediction with automated graph neural network","volume":"28","author":"Gao","year":"2024","journal-title":"IEEE J. Biomed. Health Inform"},{"key":"B11","doi-asserted-by":"crossref","first-page":"1391","DOI":"10.1109\/ICCMC65190.2025.11140940","article-title":"\u201cRAG-enhanced clinical scheduling system with named entity-aware patient workflow optimization,\u201d","volume-title":"2025 8th International Conference on Computing Methodologies and Communication (ICCMC)","author":"Gowtham","year":"2025"},{"key":"B12","doi-asserted-by":"publisher","first-page":"585","DOI":"10.1093\/jamia\/ocv153","article-title":"OpenFDA: an innovative platform providing access to a wealth of FDA's publicly available data","volume":"23","author":"Kass-Hout","year":"2016","journal-title":"J. Am. Med. Inf. Assoc."},{"key":"B13","doi-asserted-by":"publisher","first-page":"31","DOI":"10.1016\/j.seizure.2026.01.011","article-title":"Clinical reliability and concordance of drug-drug interaction tools as clinical decision support systems in patients with epilepsy","volume":"136","author":"Krayem","year":"2026","journal-title":"Seizure"},{"key":"B14","doi-asserted-by":"publisher","first-page":"D1075","DOI":"10.1093\/nar\/gkv1075","article-title":"The SIDER database of drugs and side effects","volume":"44","author":"Kuhn","year":"2016","journal-title":"Nucleic Acids Res"},{"key":"B15","doi-asserted-by":"crossref","first-page":"42","DOI":"10.1109\/BIBM62325.2024.10822490","article-title":"\u201cViDTA: enhanced drug-target affinity prediction via virtual graph nodes and attention-based feature fusion,\u201d","volume-title":"2024 IEEE international conference on Bioinformatics and Biomedicine (BIBM)","author":"Li","year":"2024"},{"key":"B16","doi-asserted-by":"crossref","first-page":"713","DOI":"10.1109\/BIBM62325.2024.10821781","article-title":"\u201cMambaDTA: a novel drug-target affinities prediction method with state-space model,\u201d","volume-title":"2024 IEEE international conference on Bioinformatics and Biomedicine (BIBM)","author":"Liu","year":"2024"},{"key":"B17","doi-asserted-by":"crossref","first-page":"64","DOI":"10.18653\/v1\/2024.knowllm-1.6","article-title":"\u201cClinicalRAG: enhancing clinical decision support through heterogeneous knowledge retrieval,\u201d","volume-title":"Proceedings of the 1st Workshop on Towards Knowledgeable Language Models (KnowLLM 2024)","author":"Lu","year":"2024"},{"key":"B18","doi-asserted-by":"publisher","first-page":"6915","DOI":"10.1109\/JBHI.2025.3561009","article-title":"A drug-drug interaction prediction method based on atomic 3D position encoding and elastic message passing graph neural network.","volume":"29","author":"Luo","year":"2025","journal-title":"IEEE J. Biomed. Health Inform"},{"key":"B19","first-page":"102","article-title":"\u201cKnowledge graph completion to predict polypharmacy side effects,\u201d","volume-title":"Proceedings of the 13th international conference on Data Integration in the Life Sciences (DILS)","author":"Malone","year":"2018"},{"key":"B20","doi-asserted-by":"publisher","first-page":"11620","DOI":"10.48550\/arXiv.2112.07837","article-title":"Central-smoothing hypergraph neural networks for predicting drug\u2013drug interactions","volume":"35","author":"Nguyen","year":"2024","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"B21","first-page":"809","article-title":"\u201cA three-way model for collective learning on multi-relational data,\u201d","volume-title":"Proceedings of the 28th International Conference on Machine Learning (ICML)","author":"Nickel","year":"2011"},{"key":"B22","doi-asserted-by":"publisher","first-page":"4","DOI":"10.3390\/ph14010004","article-title":"Improved detection criteria for detecting drug-drug interaction signals using the proportional reporting ratio","volume":"14","author":"Noguchi","year":"2021","journal-title":"Pharmaceuticals"},{"key":"B23","doi-asserted-by":"crossref","first-page":"701","DOI":"10.1145\/2623330.2623732","article-title":"\u201cDeepwalk: online learning of social representations,\u201d","volume-title":"Proceedings of the 20th ACM SIGKDD international conference on Knowledge Discovery and Data Mining (KDD)","author":"Perozzi","year":"2014"},{"key":"B24","first-page":"1029","article-title":"\u201cSpartan: scalable parafac2 for large and sparse data,\u201d","author":"Perros","year":"2017","journal-title":"Proceedings of the 23rd ACM SIGKDD international conference on Knowledge Discovery and Data Mining (KDD"},{"key":"B25","doi-asserted-by":"publisher","first-page":"48","DOI":"10.1038\/s41540-024-00374-0","article-title":"DeepARV: ensemble deep learning to predict drug-drug interaction of clinical relevance with antiretroviral therapy","volume":"10","author":"Pham","year":"2024","journal-title":"npj Syst. Biol. Appl."},{"key":"B26","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1016\/j.ijmedinf.2019.04.006","article-title":"Preventing potential drug-drug interactions through alerting decision support systems: a clinical context based methodology","volume":"127","author":"Pirnejad","year":"2019","journal-title":"Int. J. Med. Inf."},{"key":"B27","doi-asserted-by":"publisher","first-page":"1937","DOI":"10.1109\/JBHI.2013.2281505","article-title":"A novel semisupervised algorithm for rare prescription side effect discovery","volume":"18","author":"Reps","year":"2014","journal-title":"IEEE J. Biomed. Health Inf."},{"key":"B28","doi-asserted-by":"publisher","first-page":"519","DOI":"10.1002\/pds.1001","article-title":"The reporting odds ratio and its advantages over the proportional reporting ratio","volume":"13","author":"Rothman","year":"2004","journal-title":"Pharmacoepidemiol. Drug Saf"},{"key":"B29","doi-asserted-by":"publisher","DOI":"10.48550\/arxiv.2506.10896","article-title":"BioClinical ModernBERT: a state-of-the-art long-context encoder for biomedical and clinical NLP","author":"Sounack","year":"2025","journal-title":"arXiv"},{"key":"B30","doi-asserted-by":"publisher","first-page":"125r","DOI":"10.1126\/scitranslmed.3003377","article-title":"Data-driven prediction of drug effects and interactions","volume":"4","author":"Tatonetti","year":"2012","journal-title":"Sci. Transl. Med"},{"key":"B31","unstructured":"\u201cComplex embeddings for simple link prediction,\u201d\n          \n          2071\n          2080\n          \n            \n              Trouillon\n              T.\n            \n            \n              Welbl\n              J.\n            \n            \n              Riedel\n              S.\n            \n            \n              Gaussier\n              \u00c9.\n            \n            \n              Bouchard\n              G.\n            \n          \n          New York, NY\n          PMLR\n          Proceedings of the 33rd International Conference on Machine Learning (ICML)\n          \n          2016"},{"key":"B32","unstructured":"FDA Adverse Event Reporting System (FAERS) Database"},{"key":"B33","doi-asserted-by":"publisher","first-page":"146","DOI":"10.1145\/3326362","article-title":"Dynamic graph CNN for learning on point clouds","volume":"38","author":"Wang","year":"2019","journal-title":"ACM Trans. Graph"},{"key":"B34","doi-asserted-by":"publisher","first-page":"168","DOI":"10.1109\/TCBB.2020.2988018","article-title":"Predicting drug-drug interactions based on integrated similarity and semi-supervised learning","volume":"19","author":"Yan","year":"2022","journal-title":"IEEE\/ACM Trans. Comput. Biol. Bioinform"},{"key":"B35","article-title":"\u201cEmbedding entities and relations for learning and inference in knowledge bases,\u201d","volume-title":"Proceedings of the 3rd International Conference on Learning Representations (ICLR)","author":"Yang","year":"2015"},{"key":"B36","doi-asserted-by":"publisher","first-page":"i457","DOI":"10.1093\/bioinformatics\/bty294","article-title":"Modeling polypharmacy side effects with graph convolutional networks","volume":"34","author":"Zitnik","year":"2018","journal-title":"Bioinformatics"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1874825\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,29]],"date-time":"2026-07-29T06:02:23Z","timestamp":1785304943000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1874825\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,29]]},"references-count":36,"alternative-id":["10.3389\/frai.2026.1874825"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1874825","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,29]]},"article-number":"1874825"}}